Summary
In the present paper we give an evaluation of ε-entropy of a one dimensional diffusion process ξ(t) 0≦t≦T whose generator is
where the diffusion coefficienta(x) 2 satisfies
for everyx andy (L, k andK are positive constants). We assume ξ(0)=0 for simplicity of calculation. Then we can prove the following:Theorem.Under the conditions (2)and (3),ε-entropy H ε ({ξ(t)})of ξ(t), 0≦t≦T is asymptotically evaluated for small ε>0
Previously Kolmogorov stated in [3] without proof “For a diffusion process whose generator is given by (1)H ε ({ξ(t)}) is calculated by the formula:
under certain natural conditions”. However, in consideration of Pinsker’s results for Gaussian processes [5] and our present theorem this formula appears inaccurate. For the proof of the theorem we use the well-known formula of ε-entropy for finite dimensional random variables (lemma 3).
Similar content being viewed by others
References
R. Courant and D. Hilbert,Methoden der matematischen Physik I, Springer, 1931.
A. M. Geerish and P. M. Schultheiss, “Information rates of non-Gaussian processes,”IEEE Transac. Inform, Theory, IT-10 (1964), 265–271.
A. N. Kolmogorov, “Theory of transmission of information,”Amer. Math. Soc. Transl., Ser. 2, 291–321.
Yu. N. Linikov, “Calculation of ε-entropy of random variables for small ε,” (in Russian),Problems of Transmission of Information, 1, No. 2 (1965), 18–27.
M. S. Pinsker, “Gaussian sources,” (in Russian),Problems of Transmission of Information 14 (1963), 59–100.
Author information
Authors and Affiliations
Additional information
\(f(\varepsilon ) \precsim g(\varepsilon )\) means\(\overline {\mathop {\lim }\limits_{\varepsilon \to 0} } \frac{{f(\varepsilon )}}{{g(\varepsilon )}} \leqq 1\)
The Institute of Statistical Mathematics
About this article
Cite this article
Kazi, K. On the ε-entropy of diffusion processes. Ann Inst Stat Math 21, 347–356 (1969). https://doi.org/10.1007/BF02532263
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF02532263